Motion Vector Predictor Derivation with Redundancy Check
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In video coding systems, the existing Motion Vector Prediction (MVP) techniques, such as those in HEVC Test Model version 3.0 (HM-3.0), do not fully leverage redundancy checks to improve coding efficiency, as derived MVPs often lack uniqueness, leading to suboptimal bitrate reduction, especially in low-bitrate applications.
Innovation Solution
The method involves deriving motion vector predictors (MVPs) for current blocks in Inter, Skip, and Merge modes by receiving motion vectors from neighboring reference blocks, performing spatial and temporal searches with enhanced redundancy checks to ensure non-redundant MVPs are obtained, and configuring temporal search sets with prioritized motion vectors to provide unique predictors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If redundancy checking is performed during MVP derivation, then coding efficiency is improved, but computational complexity increases
Solution Approach 1:
The patent performs redundancy checking during the MVP derivation process itself, rather than after all MVPs are generated. By checking for redundancy as each spatial MVP is derived and immediately eliminating duplicate candidates, the method prevents redundant computations in subsequent steps while maintaining coding efficiency through optimized candidate selection
Solution Approach 2:
The patent extracts and removes redundant MVP candidates from the candidate list during the derivation process. By identifying and eliminating duplicate motion vector predictors as they are generated, the method reduces the number of candidates that need to be processed further, thereby improving coding efficiency without proportionally increasing computational complexity
2Adaptability or versatility
If multiple spatial MVPs are derived from neighboring blocks, then the possibility of obtaining a unique predictor increases, but redundancy increases
Solution Approach 1:
The patent implements a feedback mechanism where each derived spatial MVP is immediately checked against previously derived MVPs to detect redundancy. This real-time feedback allows the system to maintain a diverse set of predictors from multiple neighboring blocks while actively removing duplicates, thus preserving predictor availability while eliminating redundancy
Solution Approach 2:
The patent extracts and removes redundant motion vector predictors from the candidate set during derivation. By systematically eliminating duplicate MVPs while retaining unique candidates from different spatial locations, the method ensures a diverse predictor pool without wasteful redundancy, balancing adaptability and information efficiency
3Reliability
If redundancy checking is extended to temporal MVP derivation, then non-redundant MVPs are ensured, but processing time increases
Solution Approach 1:
The patent performs redundancy checking during the temporal MVP derivation process itself, integrating the check into the derivation workflow rather than as a separate post-processing step. By checking for redundancy as temporal MVPs are generated and immediately eliminating duplicates, the method ensures MVP uniqueness while minimizing additional processing time through optimized candidate evaluation
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An apparatus and method for deriving a motion vector predictor are disclosed. A search set comprising of multiple (spatial or temporal) search MVs with priority is determined, wherein the search MVs for multiple neighboring reference block or one or more co-located reference blocks are configured into multiple search MV groups. In order to improve coding efficiency, embodiments according to the present invention perform redundancy check every time after a search MV group is searched to determine whether an available search MV is found. If an available search MV is found and the available search MV is not the same as a previously derived motion vector predictor (MVP), the available search MV is used as the MVP and the MVP derivation process terminates. Otherwise, the MVP derivation process moves to the next reference block. The search MV group can be configured to include different search MV(s) associated with reference blocks.